package main
import (
- "flag"
"fmt"
"html"
"io"
- "mime"
"net/http"
"net/url"
"os"
- "regexp"
+ "path"
"strconv"
"strings"
+ "sync"
+ "time"
"git.curoverse.com/arvados.git/sdk/go/arvadosclient"
"git.curoverse.com/arvados.git/sdk/go/auth"
"git.curoverse.com/arvados.git/sdk/go/keepclient"
)
-type handler struct{}
-
-var (
- clientPool = arvadosclient.MakeClientPool()
- trustAllContent = false
- attachmentOnlyHost = ""
-)
-
-func init() {
- flag.StringVar(&attachmentOnlyHost, "attachment-only-host", "",
- "Accept credentials, and add \"Content-Disposition: attachment\" response headers, for requests at this hostname:port. Prohibiting inline display makes it possible to serve untrusted and non-public content from a single origin, i.e., without wildcard DNS or SSL.")
- flag.BoolVar(&trustAllContent, "trust-all-content", false,
- "Serve non-public content from a single origin. Dangerous: read docs before using!")
+type handler struct {
+ Config *Config
+ clientPool *arvadosclient.ClientPool
+ setupOnce sync.Once
}
-// return a UUID or PDH if s begins with a UUID or URL-encoded PDH;
-// otherwise return "".
+// parseCollectionIDFromDNSName returns a UUID or PDH if s begins with
+// a UUID or URL-encoded PDH; otherwise "".
func parseCollectionIDFromDNSName(s string) string {
// Strip domain.
if i := strings.IndexRune(s, '.'); i >= 0 {
var urlPDHDecoder = strings.NewReplacer(" ", "+", "-", "+")
-// return a UUID or PDH if s is a UUID or a PDH (even if it is a PDH
-// with "+" replaced by " " or "-"); otherwise return "".
+// parseCollectionIDFromURL returns a UUID or PDH if s is a UUID or a
+// PDH (even if it is a PDH with "+" replaced by " " or "-");
+// otherwise "".
func parseCollectionIDFromURL(s string) string {
if arvadosclient.UUIDMatch(s) {
return s
return ""
}
+func (h *handler) setup() {
+ h.clientPool = arvadosclient.MakeClientPool()
+}
+
+// ServeHTTP implements http.Handler.
func (h *handler) ServeHTTP(wOrig http.ResponseWriter, r *http.Request) {
+ h.setupOnce.Do(h.setup)
+
var statusCode = 0
var statusText string
httpserver.Log(remoteAddr, statusCode, statusText, w.WroteBodyBytes(), r.Method, r.Host, r.URL.Path, r.URL.RawQuery)
}()
+ if r.Method == "OPTIONS" {
+ method := r.Header.Get("Access-Control-Request-Method")
+ if method != "GET" && method != "POST" {
+ statusCode = http.StatusMethodNotAllowed
+ return
+ }
+ w.Header().Set("Access-Control-Allow-Headers", "Range")
+ w.Header().Set("Access-Control-Allow-Methods", "GET, POST")
+ w.Header().Set("Access-Control-Allow-Origin", "*")
+ w.Header().Set("Access-Control-Max-Age", "86400")
+ statusCode = http.StatusOK
+ return
+ }
+
if r.Method != "GET" && r.Method != "POST" {
statusCode, statusText = http.StatusMethodNotAllowed, r.Method
return
}
- arv := clientPool.Get()
+ if r.Header.Get("Origin") != "" {
+ // Allow simple cross-origin requests without user
+ // credentials ("user credentials" as defined by CORS,
+ // i.e., cookies, HTTP authentication, and client-side
+ // SSL certificates. See
+ // http://www.w3.org/TR/cors/#user-credentials).
+ w.Header().Set("Access-Control-Allow-Origin", "*")
+ }
+
+ arv := h.clientPool.Get()
if arv == nil {
- statusCode, statusText = http.StatusInternalServerError, "Pool failed: "+clientPool.Err().Error()
+ statusCode, statusText = http.StatusInternalServerError, "Pool failed: "+h.clientPool.Err().Error()
return
}
- defer clientPool.Put(arv)
+ defer h.clientPool.Put(arv)
pathParts := strings.Split(r.URL.Path[1:], "/")
var reqTokens []string
var pathToken bool
var attachment bool
- credentialsOK := trustAllContent
+ credentialsOK := h.Config.TrustAllContent
- if r.Host != "" && r.Host == attachmentOnlyHost {
+ if r.Host != "" && r.Host == h.Config.AttachmentOnlyHost {
credentialsOK = true
attachment = true
} else if r.FormValue("disposition") == "attachment" {
} else if len(pathParts) >= 3 && pathParts[0] == "collections" {
if len(pathParts) >= 5 && pathParts[1] == "download" {
// /collections/download/ID/TOKEN/PATH...
- targetID = pathParts[2]
+ targetID = parseCollectionIDFromURL(pathParts[2])
tokens = []string{pathParts[3]}
targetPath = pathParts[4:]
pathToken = true
} else {
// /collections/ID/PATH...
- targetID = pathParts[1]
- tokens = anonymousTokens
+ targetID = parseCollectionIDFromURL(pathParts[1])
+ tokens = h.Config.AnonymousTokens
targetPath = pathParts[2:]
}
- } else {
+ }
+
+ if targetID == "" {
statusCode = http.StatusNotFound
return
}
- if t := r.FormValue("api_token"); t != "" {
+
+ formToken := r.FormValue("api_token")
+ if formToken != "" && r.Header.Get("Origin") != "" && attachment && r.URL.Query().Get("api_token") == "" {
+ // The client provided an explicit token in the POST
+ // body. The Origin header indicates this *might* be
+ // an AJAX request, in which case redirect-with-cookie
+ // won't work: we should just serve the content in the
+ // POST response. This is safe because:
+ //
+ // * We're supplying an attachment, not inline
+ // content, so we don't need to convert the POST to
+ // a GET and avoid the "really resubmit form?"
+ // problem.
+ //
+ // * The token isn't embedded in the URL, so we don't
+ // need to worry about bookmarks and copy/paste.
+ tokens = append(tokens, formToken)
+ } else if formToken != "" {
// The client provided an explicit token in the query
// string, or a form in POST body. We must put the
// token in an HttpOnly cookie, and redirect to the
// It is not safe to copy the provided token
// into a cookie unless the current vhost
// (origin) serves only a single collection or
- // we are in trustAllContent mode.
+ // we are in TrustAllContent mode.
statusCode = http.StatusBadRequest
return
}
http.SetCookie(w, &http.Cookie{
Name: "arvados_api_token",
- Value: auth.EncodeTokenCookie([]byte(t)),
+ Value: auth.EncodeTokenCookie([]byte(formToken)),
Path: "/",
HttpOnly: true,
})
if credentialsOK {
reqTokens = auth.NewCredentialsFromHTTPRequest(r).Tokens
}
- tokens = append(reqTokens, anonymousTokens...)
+ tokens = append(reqTokens, h.Config.AnonymousTokens...)
}
if len(targetPath) > 0 && targetPath[0] == "_" {
statusCode, statusText = http.StatusInternalServerError, err.Error()
return
}
+ if kc.Client != nil && kc.Client.Transport != nil {
+ // Workaround for https://dev.arvados.org/issues/9005
+ if t, ok := kc.Client.Transport.(*http.Transport); ok {
+ defer t.CloseIdleConnections()
+ }
+ }
rdr, err := kc.CollectionFileReader(collection, filename)
if os.IsNotExist(err) {
statusCode = http.StatusNotFound
}
defer rdr.Close()
- basenamePos := strings.LastIndex(filename, "/")
- if basenamePos < 0 {
- basenamePos = 0
- }
- extPos := strings.LastIndex(filename, ".")
- if extPos > basenamePos {
- // Now extPos is safely >= 0.
- if t := mime.TypeByExtension(filename[extPos:]); t != "" {
- w.Header().Set("Content-Type", t)
- }
- }
- if rdr, ok := rdr.(keepclient.ReadCloserWithLen); ok {
- w.Header().Set("Content-Length", fmt.Sprintf("%d", rdr.Len()))
- }
-
- applyContentDispositionHdr(w, r, filename[basenamePos:], attachment)
- rangeRdr, statusCode := applyRangeHdr(w, r, rdr)
+ basename := path.Base(filename)
+ applyContentDispositionHdr(w, r, basename, attachment)
- w.WriteHeader(statusCode)
- _, err = io.Copy(w, rangeRdr)
+ modstr, _ := collection["modified_at"].(string)
+ modtime, err := time.Parse(time.RFC3339Nano, modstr)
if err != nil {
- statusCode, statusText = http.StatusBadGateway, err.Error()
- }
-}
-
-var rangeRe = regexp.MustCompile(`^bytes=0-([0-9]*)$`)
-
-func applyRangeHdr(w http.ResponseWriter, r *http.Request, rdr keepclient.ReadCloserWithLen) (io.Reader, int) {
- w.Header().Set("Accept-Ranges", "bytes")
- hdr := r.Header.Get("Range")
- fields := rangeRe.FindStringSubmatch(hdr)
- if fields == nil {
- return rdr, http.StatusOK
- }
- rangeEnd, err := strconv.ParseInt(fields[1], 10, 64)
- if err != nil {
- // Empty or too big for int64 == send entire content
- return rdr, http.StatusOK
- }
- if uint64(rangeEnd) >= rdr.Len() {
- return rdr, http.StatusOK
+ modtime = time.Now()
}
- w.Header().Set("Content-Length", fmt.Sprintf("%d", rangeEnd+1))
- w.Header().Set("Content-Range", fmt.Sprintf("bytes %d-%d/%d", 0, rangeEnd, rdr.Len()))
- return &io.LimitedReader{R: rdr, N: rangeEnd + 1}, http.StatusPartialContent
+ http.ServeContent(w, r, basename, modtime, rdr)
}
func applyContentDispositionHdr(w http.ResponseWriter, r *http.Request, filename string, isAttachment bool) {